Model workspace

AI video models

Choose the right model workspace for image-to-video generation, cinematic motion, and production-ready video drafts.

Nano Banana 2

Use Nano Banana for fast, low-cost drafts or choose Nano Banana 2 for clearer text, up to 4K output, and multi-image editing.

Open model

Veo 3.1 cinematic video

Veo 3.1 is tuned for cinematic image-to-video and text-to-video workflows with polished motion and strong visual continuity.

Open model

Seedance 2.5 image to video

Seedance 2.0 focuses on fast, flexible image-to-video generation with aspect ratio, resolution, and duration controls.

Open model

MiniMax H3 AI video

Create cinematic clips with MiniMax H3, a multimodal AI video model for text, image, video, and audio-guided generation.

Open model

Kling 3.0 AI video

Create from a prompt or reference image with the live Kling model family and compare generation modes in one workspace.

Open model

Wan 3.0

Create with Wan 3.0 and explore the Wan model family in one workspace designed for image-led AI video production.

Open model

How to choose your image-to-video model

Compare each live model workspace by starting input, controls, pacing, and creative goal. Choose the family that best matches the shot you need to produce.

Choose Seedance 2.0 for control and speed

Seedance 2.0 is built for repeatable product and commerce work: first-frame and first-and-last-frame modes, selectable duration, resolution, and aspect ratio, plus fast variants for cheap prompt testing across formats.

Choose Veo 3.1 for cinematic motion

Veo 3.1 leans into camera language and atmosphere, with polished motion and strong visual continuity for launches, brand storytelling, and prompt-led scenes.

Choose Hailuo for image-led motion

MiniMax H3 brings 2K output, 5–15-second clips, multimodal references, first-and-last-frame control, and video editing into one creative model.

Choose Kling for a flexible model family

Move between current Kling versions as the brief changes, from prompt-led exploration to first-frame animation and guided transitions.

One workspace for every model

Switch models without leaving the workspace. Start from a reference image or a prompt, keep your library and history in one place, and move a project between models as the creative direction changes.